Horses represent a natural host for alphaherpesviruses (EHV-1 and EHV-4), and gammaherpesviruses (EHV-2 and EHV-5) [2, 16], with EHV-1 being the most common viral cause of abortion storms in horses.
Usually, these four EHVs occur together in aborted samples [2, 16] and hence, the findings of this study are in agreement with published reports regarding the detection of EHV-2, EHV-4, and EHV-5 as co-infections in aborted fetuses [2]. However, although coinfections with EHV-2 (accession nos: OK362329-OK362331), EHV-4 (accession no: OK362332), EHV-5 (accession nos: OL440383-OL440384) alongside EHV-1 were detected, only the latter was identified in all the fetal samples of the unvaccinated Yili horses. Therefore, the results suggested that EHV-1 could have been the main cause of the abortion storm in the Yili horse population included in this study. In fact, this is supported by previous results which revealed that EHV-1 was frequently involved in outbreaks that led to horse abortions around the world [2, 11,12,13,14,15,16,17,18,19,20,21,22,23,24]. Regarding the other viruses, EHV-4 is known to provoke sporadic abortions, but it has not yet been demonstrated whether EHV-2 and EHV-5 are also a cause for concern [16, 25, 26]. Given that these three viruses were detected in the EHV-1 positive aborted samples, there is the need to isolate them in order to demonstrate whether they can be the cause of abortion in Yili horses. In this context, a previous study actually noted that these three viruses circulated in thoroughbred foals with respiratory diseases in North Xinjiang [27], but here, the detection of these strains is being reported for the first time in abortion cases of horses in North Xinjiang. Therefore, based on the results, it can be concluded that EHV-2, EHV-4 and EHV-5 might contribute to abortions and respiratory diseases in the horses of Xinjiang, China.
In a similar way to one previous study, some variations in ORF33 genes of the EHV-1 viruses were also detected in the aborted samples [11]. Indeed, the partial ORF33 (592 nt) sequences from the 43 EHV-1 shared 99.3–100% nt identity, and clustered into six groups, thereby indicating genetic diversity.
Previous studies have shown that the nucleotide at position 2254 of ORF30 was strongly associated with the occurrence of EHV-1 neurological diseases [11,12,13,14,15,16,17,18,19,20,21,22,23,24]. For instance, most of the viruses recovered from abortion cases worldwide contained adenine (A) at this position (asparagine (N) at position 752 of the protein) and belonged to non-neuropathogenic strains (associated with non-neurological conditions) [11,12,13,14,15,16,17,18,19,20,21,22,23,24]. Similarly, as reported by Smith et al. (2010), G2254 strains have been isolated less frequently than A2254 ones in abortion cases [15]. However, in this study, the ORF30 sequences of 43 EHV-1 detected in the present study carried a G at position 2254 (D in position 752) (Fig. 2). Hence, the data provided evidence that neuropathogenic EHV-1 could have been responsible for the series of abortions in Yili horses. It should nevertheless be noted that comparative sequence analysis also showed similarity to EHV-1 YM2019 (accession no. MT063054) which, according to GenBank information, is non-neuropathogenic and was identified in aborted samples of Przewalski’s horses. As such, the results indicated that both non-neuropathogenic and neuropathogenic EHV-1 strains could be circulating in the horse population of Xinjiang, China. Indeed, while non-neuropathogenic EHV-1 had been detected from some cases of neurological diseases, in most cases, neuropathogenic EHV-1 was associated with the development of neurological diseases, higher levels of viraemia as well as longer disease duration. For instance, in Argentina, the United Kingdom and the United States, approximately 8.7–50% of neuropathogenic EHV-1, identified from abortion cases, were associated with neurological disease [12, 14]. These data clearly supported the results regarding the presence of neuropathogenic EHV-1 in Yili horses despite the absence of neurological signs.
Several published studies showed that only non-neuropathogenic EHV-1 are isolated in Brazil, Turkey and Poland [13, 17, 18]. Neuropathogenic EHV-1 has a low prevalence in Japan (2.7%), the United States (10.8–19.4%), Argentina (7.4%), France (24%), and Germany (10.6%) [4, 14, 15, 19,20,21], but a high prevalence in Italy (90%), Uruguay (92.3%) and Ethiopia (98.9%) [11, 22, 23]. However, over the past few decades, two investigations conducted in Italy and the United States indicated that the prevalence of neuropathogenic EHV-1 was on the rise. In addition, even though vaccination could reduce virus spreading, one study in Germany reported that vaccination against EHV-1 was not always successful at preventing the spread of the neuropathogenic genotype [24], with another study in Italy confirming that the spread of neuropathogenic EHV-1 strains were could be observed in both vaccinated mares as well as unvaccinated horses [11]. Thus, it is not surprising that all EHV-1 detected in aborted samples from unvaccinated Yili horses, could have been neuropathogenic EHV-1.
Furthermore, two recent studies in France and the United States identified a new EHV-1 DNA polymerase (ORF30) genotype. This was followed by the isolation of a C2254/H752 which was then associated with both respiratory and neurological clinical signs [28, 29]. Further study will investigate the presence of this new EHV-1 genotype in China’s horses.